Table 3

Advantages and disadvantages of computational and subproteomic approaches to localization analysis.

Computational methods

Proteomics analysis

Advantages

Rapid predictions for all proteins deduced to be encoded in a given sequence

Can be performed under different conditions and provide condition-specific information

Detailed information about specific features of proteins, e.g. signal peptides, TMHs

Confirms expression of hypothetical proteins

Identification of potential contaminants in subproteome analyses

Large-scale source of data on SCL for hypothetical proteins that cannot be easily predicted computationally

Identification of hydrophobic integral membrane proteins

Disadvantages

Does not perform as well (less predictions) when analyzing an organism that is not similar to well studied/model organisms.

Time-consuming

May miss flagging some multiply-localized proteins

Low abundance and hydrophobic proteins not readily detected

Poorly predicts particular localizations for which there is little training data, or the proteins are computationally difficult to differentiate between localizations.

Difficult to accurately identify all proteins found on the gel

Cannot identify condition-specific data on SCL, particularly proteins that change SCL depending on the condition.

One subcellular fraction at once analyzed

Subfractionation often results in contamination

Cannot identify multiply localized proteins

Rey et al. BMC Genomics 2005 6:162   doi:10.1186/1471-2164-6-162